Multifaceted radiation detection and classification system
A system identifying a source of radiation is provided. The system includes a radiation source detector and a radiation source identifier. The radiation source detector receives measurements of radiation; for one or more sources, generates a detection metric indicating whether that source is present in the measurements; and evaluates the detection metrics to detect whether a source is present in the measurements. When the presence of a source in the measurements is detected, the radiation source identifier for one or more sources, generates an identification metric indicating whether that source is present in the measurements; generates a null-hypothesis metric indicating whether no source is present in the measurements; evaluates the one or more identification metrics and the null-hypothesis metric to identify the source, if any, that is present in the measurements.
1. A method for a radiation source detector to generate a metric relating to a source of radiation in measurements of radiation, the method comprising:
accessing, by the radiation source detector, a measurement, a source signature of the source and an estimated background;
generating, by the radiation source detector, a background vector factoring out the source signature;
generating, by the radiation source detector, a projection vector based on a weighting matrix derived from the estimated background and based on the source signature and the estimated background;
generating, by the radiation source detector, the metric based on the measurement, the projection vector, and an expected variance and an average background; and
when the metric satisfies a threshold, outputting, by the radiation source detector, an indication that a presence of the source has been detected.
2. The method of claim 1 , wherein the source signature is a histogram representing an energy range divided into energy bins, each energy bin having a value representing a count of photons emitted by that source over a time interval.
3. The method of claim 2 , wherein the weighting matrix is generated from the estimated background wherein a variance of each energy bin is based on the expected background.
4. The method of claim 1 , wherein a source signature represents a shielding of the source.
5. The method of claim 1 , further comprising generating aggregated measurements of different numbers of measurements, and for each aggregated measurement, generating a metric based on the aggregated measurement and the projection vector.
6. The method of claim 1 , wherein the indication that the presence of the source has been detected comprises an indication that the source has been semi-definitively detected.
7. The method of claim 1 , wherein the projection vector is generated according to the equation,
T=kS t W ( I−{circumflex over (B)} ( {circumflex over (B)} t W{circumflex over (B)} ) −1 {circumflex over (B)} t W ),
where T represents the projection vector, S represents the source signature, W represents basis vectors, {circumflex over (B)} is represented by the equation,
{circumflex over (B)} =( I−μS ( S t S ) −1 S ),
and k and μ represents variables adapted to produce a metric with a unit variance with μ represented by the equation,
μ=T B ,
where B represents average background;
and wherein the metric is generated according to the equation,
DM
=
TX
-
μΔ
t
❘
"\[LeftBracketingBar]"
B
Δ
t
❘
"\[RightBracketingBar]"
1
¶
,
where DM represents the metric, X represents the measurement, t represents time of the measurement, Δt represents change in time period, and B represents expected background at time t.
8. The method of claim 1 , wherein the method comprises solving the equation,
[
S
B
0
λ
1
_
]
[
s
b
_
]
=
[
X
λ
e
b
]
,
where S represents the source signature, 1 represents a row vector of ones equal to a number of background components, s represents estimated counts in the source, b represents a vector of background intensities in each component, B is a matrix of background basis vectors, X is measurement, λ is an estimated total count in the background, A is a tuning parameter based on reliability of the estimated background total counts;
and wherein the metric is generated according to the equation,
DM
=
s
-
κmax
(
b
-
e
b
,
0
)
e
b
,
where DM represents the metric and K represents a penalty for any difference in the estimated total count in the background and an expected background total count in the background.
9. One or more computing systems of a radiation source detector to generate a metric relating to a source of radiation in measurements of radiation, the one or more computing systems comprising:
one or more computer-readable storage mediums storing computer-executable instructions, execution of which causes the one or more computing systems to:
access a source signature of the source and an estimated background;
generate a background vector factoring out the source signature;
generate a projection vector based on a weighting matrix derived from the estimated background and based on the source signature and the estimated background;
access a measurement;
generate the metric based on the measurement, the projection vector, and an expected variance and an average background; and
when the metric satisfies a threshold, output an indication that a presence of the source has been detected; and
one or more processors for executing the computer-executable instructions stored in the one or more computer-readable storage mediums.
10. The one or more computing systems of claim 9 , wherein the source signature is a histogram representing an energy range divided into energy bins, each energy bin having a value representing a count of photons emitted by that source over a time interval.
11. The one or more computing systems of claim 10 , wherein the weighting matrix is generated from the estimated background wherein a variance of each energy bin is based on the expected background.
12. The one or more computing systems of claim 9 , wherein a source signature represents a shielding of the source.
13. The one or more computing systems of claim 9 , wherein execution of the instructions further generates aggregated measurements of different numbers of measurements and, for each aggregated measurement, generates a metric based on the aggregated measurement and the projection vector.
14. The one or more computing systems of claim 9 , wherein the indication that the presence of the source has been detected comprises an indication that presence of the source has been semi-definitively detected.
15. The one or more computing systems of claim 9 , wherein the projection vector is generated according to the equation,
T=kS t W ( I−{circumflex over (B)} ( {circumflex over (B)} t W{circumflex over (B)} ) −1 {circumflex over (B)} t W ),
where T represents the projection vector, S represents the source signature, W represents basis vectors, {circumflex over (B)}is represented by the equation,
{circumflex over (B)} =( I−μS ( S t S ) −1 S ),
and k and μ represents variables adapted to produce a metric with a unit variance with μ represented by the equation,
μ=T B ,
where B represents average background;
and wherein the metric is generated according to the equation,
DM
=
TX
-
μΔ
t
❘
"\[LeftBracketingBar]"
B
Δ
t
❘
"\[RightBracketingBar]"
1
¶
,
where DM represents the metric, X represents the measurement, t represents time of the measurement, Δt represents change in time period, and B represents expected background at time t.
16. The one or more computing systems of claim 9 , wherein execution of the computer-executable instructions further causes the one or more computing systems to solve the equation,
[
S
B
0
λ
1
_
]
[
s
b
_
]
=
[
X
λ
e
b
]
,
where S represents the source signature, 1 represents a row vector of ones equal to a number of background components, s represents estimated counts in the source, b represents a vector of background intensities in each component, B is a matrix of background basis vectors, X is measurement, e b is an estimated total count in the background, λ is a tuning parameter based on reliability of the estimated background total counts;
and wherein the metric is generated according to the equation,
DM
=
s
-
κmax
(
b
-
e
b
,
0
)
e
b
,
where DM represents the metric and κ represents a penalty for any difference in the estimated total count in the background and an expected background total count in the background.